The COVID-19 pandemic, coupled with increasingly common climate-based natural disasters, showed us how vulnerable global supply chains are. But while a broken supply chain in the automobile industry may mean a shortage of spark plugs at your local auto repair shop, the same situation in the healthcare industry can result in the inability to effectively treat illness or injury.
Many of us tend to be surprised when we realize healthcare is not immune to market forces. Medical logistics includes medical devices, from oxygen pumps to scalpels; prescription drugs and other medication; personal protective equipment (PPE); healthcare furniture; and disinfectants. These are channeled through almost 6,100 hospitals and 67,000 pharmacies handling about 6 billion prescriptions in the United States alone. Cast the definition of logistics a little wider and you have to account for the indispensable human equation: doctors, nurses, technologists, specialists, and administrators.
By innovating data sharing and collaboration with partners and agencies, we may be able to better identify and more quickly analyze the most vulnerable links of the healthcare chain, predicting supply chain shortfalls and other issues before they happen.
Looking backward, looking inward
“Historically, healthcare has not used demand forecasting systems and demand management systems that many other industries do, because of the fact that all they’ve really had in terms of data is looking backwards,” said Leisa Maddoux, Global and U.S. Health Transformation Leader at EY. “One of the things the pandemic taught us was when you have something that is unpredictable, you have to become more agile in predicting.”
Fortunately, according to Gaurav Kaushik, Founder and President of ScienceIO, AI can be of assistance, much more today than in years past. “AI is often better at any individual task when [it’s] trained to do multiple things at once,” he said. “We’ve learned a lot about how multi-task learning can improve the quality of AI.”
Additionally, medical organizations often compete with manufacturers for the same types of items. Even if those organizations are nonprofit, they are sometimes still in competition.
“You’ll have five different types of pacemaker, and so you have five different manufacturers, and they don’t want to share with each other, and they don’t want you sharing,“ said Maddoux.
It is not a matter of lacking the data, said Maddoux: “Hospitals have data, the distributors have data, the manufacturers have data.” They just have to work together to figure out how they can use it for everyone’s benefit.
Data is critical to healthcare, said Lisa Anderson, President of the LMA Consulting Group. “Data can lead to figuring out better where your product is,” she said. “It allows you to track the demand for your product, to keep critical supplies in stock where they need to be. Everyone I’ve talked to believes data sharing is the future. The problem is, it is really quite hard to get competitors to agree to share data when you don’t know if you’re going to come out on the equal side of that situation.”
There are approximately 11,000 rogue pharmacies on the internet and operating throughout the world, according to Jesse Cugliotta, Snowflake’s Global Industry GTM Lead for Healthcare and Life Sciences. “There are real safety issues associated with people thinking that they’re buying antibiotics or expensive oncology drugs, but they wind up with counterfeits whose efficacy is not up to par or which are downright dangerous,” he said.
One example of the data that is absolutely necessary to track when it comes to pharmaceuticals is the temperature needs of medicines, known as the “cold chain.” It’s a safety issue, but it also costs companies across the world $35 billion annually.
As the pandemic made clear and urgent, tracking, using, and sharing data cannot be back-burnered indefinitely.
We work together or we fall apart
To move into a profitable data-sharing mindset, one that can strengthen healthcare supply chains and make them more flexible, we need practical solutions. AI and the use of the cloud, for instance, means we can respond to changes and obstacles faster and with fewer dangers.
“Figuring out patterns in data to see what’s needed in what place is something AI can do better than a human,” said Anderson.
An impartial AI “referee” that can abide by HIPAA and GDPR privacy rules and other needs would take us a big step closer to sharing without fear. But we are far from creating an AI capable of doing something so complex without making a hash of things. So far, AI is designed to do one specific thing at a time. Perhaps we can build up a system from the specific experiments we make, for instance, using AI to track cargo on container ships.
“With data sharing, there’s a huge opportunity to improve profitability, market share, and revenue capture with data sharing,” said Cugliotta. “Because the one thing that everyone has realized is that for the last 40 years, every supply chain in every industry has been really looking at cost reduction as their number one driver for everything they do. But there’s a kind of a brittleness to those types of supply chains, even if they were super cost-effective. We can no longer afford to be blindsided. I need to see what’s happening across all of my suppliers at all times.“
One of the traditional limitations of AI has been the opaqueness of its decision-making. AI is not programmed with simple instructions. It’s trained, a process that is difficult to track. But that opacity is starting to clarify.
“We train machine learning models to do complicated things that are not easy for people to create rules around,” said Kaushik. “But there are methods to better understand how models will behave, test their biases, and see how they make decisions so they can be trained to be better.”
There are enough tools available to strengthen the healthcare supply chain even during times of stress and challenge. Given just how much change the future may hold, this realization should buttress the work of those responsible for our healthcare.
For more information on the challenges and possibilities of the healthcare supply chain, refer to the following resources
- Turning Data into Evidence: The Next-Generation ‘Blue Button’ (Snowflake)
- Key Determinants for Resilient Health Care Supply Chains (Deloitte)
- Stanford Health Care’s Innovative Supply Chain Model (Healthleaders)
- Beyond the Pandemic: Mitigating Supply Chain Risk and Disruption (American Hospital Association)
- Building Resilient, Intelligent Supply Chains (IBM)
